Binomial Distribution is a discrete distribution that occurs when there is only two possible outcomes

Coin experiment is a binomial distribution with probability 50% of Tails and 50% of Heads
We refer to an experiment as a trial. For instance the number of times we toss the coin
And we need to label one of the outcomes as a success, the other one is called failure
Let's assume that we toss the coin 30 times, and we define Tails as the successful outcome

Number of trials 30 tosses of the coin

Probability of success 50% landing tails

There are two formulas that are similar to the Normal Distribution, which must be used when we are dealing with a

Event # of successes Probability

We obtain exactly 20 Tails 20 2.8% Notice how we are using the last argument as False. Wh
We obtain no more than 20 Tails 20 97.9% Here we use the last argument as True, since we are dea
We obtain less than 20 Tails 19 95.1% Less than 20 tails means <20
At least 12 Tails 12 90.0% Here we are looking for the right hand side. In other wo
More than 12 Tails 12 81.9%

Between 12 and 20 Tails

20 97.9% First calculate the probability of 20 or less
11 10.0% Next calculate the probability of less than 11

87.8% Substract both results

# of Tails Probability of Exactly that number

0 0.0%
1 0.0%
2 0.0%
3 0.0%
4 0.0%
5 0.0%
6 0.1%
7 0.2%
8 0.5%
9 1.3%
10 2.8%
16.0%
11 5.1%
14.0%

12.0%

10.0%
 16.0%

12 8.1% 14.0%

13 11.2% 12.0%
14 13.5%
10.0%
15 14.4%
16 13.5% 87.8% TRUE 8.0%
17 11.2% 6.0%
18 8.1%
4.0%
19 5.1%
20 2.8% 2.0%
21 1.3%
0.0%
22 0.5% 0 2 4
23 0.2%
24 0.1%
25 0.0%
26 0.0%
27 0.0%
28 0.0%
29 0.0%
30 0.0%
more 0%
100.0% It must add to 100%

We can also use Norm.Inv when there is a desired probability and we want to know the # of success that would ge

Cumulative Probabilty # of Tails

10% 11
15% 12
20% 13
30% 14 Notice how the nature of the curve causes to lump together t
probability. Take a look above on the probability for each exa
35% 14 Tails or less adds up to 42.8%. Whereas the probability of obt
40% 14
e outcomes

en we are dealing with a Binomial Distribution

he last argument as False. When using the Binomial Distribution we can measure the exact are under the curve, since this is a discrete variable
ment as True, since we are dealing with a cumulative probability. Also notice that no more than 20, means <=20

e right hand side. In other words, we want 12 Tails or more. Therefore we find the probability of obtaining 9 tails or less, and substract from 1

obability of 20 or less
obability of less than 11 successes

Probability of Exactly that number

16.0%

14.0%

12.0%

10.0%
 Probability of Exactly that number
16.0%

14.0%

12.0%

10.0%

8.0%

6.0%

4.0%

2.0%

0.0%
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

of success that would get us there

auses to lump together the same # of trials with multiple values for the cumulative
probability for each exact value, and add them together. The probability of obtaining 14
eas the probability of obtaining a 13 Tails adds up to 29.2%
The formulas for the mean and the standard deviation in binomial distribution vary slightly.
This is because we are dealing with only two possible outcomes. Measuring the dispersion is different than when d

The formal notation to denote we are dealing with a binomial outcome is to use the Chi (or X) symbol.
We use the letter E to denote mean
And the formula for standard deviation becomes the square root of the number of trials * probability of success * (

n 100 trials
p 60% success
1-p (or q) 40% The complement to 100% from the success %. Also known as q

E(X) 60 If an experiment has 60% of success, and we conduct the experiment 100 times. We expect t
Stdev (X) 4.90 While this calculation is less intuitive, the interpretation is the same as with the normal distrib

Mu - 2σ Mu + 2σ
50 70 It is very likely that the actual number of experiments with an succesful outcome w
 ion is different than when dealing with all possible continous values

i (or X) symbol.

s * probability of success * (1 - probability of success)

ment 100 times. We expect to have 60 success, on average

me as with the normal distribution. We would expect for 95% of the chances of success to be within two standard deviations away from m

with an succesful outcome will be somewhere between 50 and 70

dard deviations away from mean
Airlines know that certain % of passengers will be a no-show.
Therefore, they sell more tickets than seats available to increase chances of running flights at full capacity.

An airline knows that the no-show rate for certain route is 10%.
The aircraft can carry 200 passengers
What should be the # of total tickets they should sell so that most of the time they end up with the 200 seats filled?

number of trials 200

probability of success 10% 1-p 90.0%

E(x) 20
stdev 4.242640687
Mu - 2σ 12

212 228
 at full capacity.

with the 200 seats filled?

Mu + 2σ 28